Instrumentation, Control & Industrial Practice Laboratory Manual Instrumentation, Control & Industrial Practice Laboratory Manual

INSTRUMENTATION AND CONTROL LABORATORY

Rigs 1-6 3. ORGANIZATION AND PREPARATION
You are expected to prepare for the Laboratory!
1. INTRODUCTION
Before practical work can start, the Laboratory Supervisor will want to see evidence of your
preparatory work. This can be in the form of initial experimental plan and observations collected using
The laboratory is an introduction to the dynamic behaviour of pilot-scale equipment and the
the “Virtual Process Control Rig”, summarized in your laboratory notebook.
operation of a feedback controller. The principle aim is understand the role and operation of the
main components in a feedback loop: (i) sensor, (ii) actuator (control valve) and (iii) controller.
Experimental work is performed by pairs (or groups of three) of investigators working on a
specific rig in the Control Laboratory. Each pair of investigators is responsible for starting their
The plant to be controlled is a liquid surge tank with a pump in the outlet stream. To an
apparatus at the beginning of the laboratory and shutting down the equipment before leaving the
acceptable approximation (which will be tested in the laboratory), the liquid flow from the tank is
laboratory. The nominal procedures in Appendix A can be adapted to the characteristics of each
independent of the liquid level in the tank. The flow of liquid entering the tank is independent of
individual rig.
the tank's liquid level; as a consequence, the tank acts as an integrator, integrating the difference
between the flow in and the flow out. Control action is needed to stabilize the liquid level by
making one of the flows dependent on the liquid level.
ƒ Before the timetabled practical session:

1. prepare a laboratory notebook where you will record all observations during
preparation and the experiments
2. read these notes carefully
3. download from the LearnServer the installation kit of the “Virtual Process Control
Laboratory” and familiarize yourself with the software and procedure.

ƒ Preparation in the lab

4. Trace all lines (pneumatic, electrical, hydraulic) on the rig.

5. Before turning on the input flowrate, observe the action of the control valve. Set the
control mode to MANUAL on the software interface and adjust the input signal to the
Figure 1. Simplified Diagram of a Surge Tank Control System valve between 0 and 100%. Record whether the valve opens or closes when you
increase/decrease the signal.
2. OBJECTIVES 6. Test the approximation that the output flow is independent of the water level in the
tank. Set the control valve on MANUAL mode, completely open and adjust the input
A. Week 1 flow rate to achieve a steady low level in the tank. Record the output flow rate. Increase
the input flow rate and adjust it to achieve a steady high level. Record the output flow
A1. Calibrate the level sensor. rate in this case. Is there a difference between the maximum output flowrates in the two
A2. Calibrate the control valve. Study the installed characteristics of a valve. cases? After this procedure establish a steady state at about 30-50 % level and record
A3. Study the effect of hysteresis in the control valve and sensor. the steady state operating conditions in your laboratory notebook.
7. Using a computer drafting package, construct a piping and instrumentation diagram
B. Week 2 (P&ID) for the experimental rig. Label all manual valves.
8. Develop an experimental procedure.
B1. Understand the dynamic behaviour of the system
B2. Evaluate manual control.
B3. Evaluate the tuning of a PID control via manual and automatic tuning
B4. Evaluate the effect of positive or negative feedback on the closed-loop system
ZK Nagy 1 ZK Nagy 2
Instrumentation, Control & Industrial Practice Laboratory Manual Instrumentation, Control & Industrial Practice Laboratory Manual

4. EXPERIMENTAL PROCEDURE 4. Plot the calibration curve (level versus voltage) and determine the gain and zero of the
instrument by fitting a line to the experimental data. Perform separate fittings of the
increasing and deceasing data, as well as using all data points. Compare the three
I. Sensor Calibration
calibration equations with each other and with the calibration obtained using the data at
point 2.
Introduction

A level sensor is an essential element in all chemical plants. The signal obtained from the device
is usually a standard electric signal (e.g. 0-5 V). This value needs to be correlated with the real
level in the tank to obtain the desired information. This process is called calibration. For II. Calibration and Characteristics of the Control Valve
appropriate level control and monitoring the characteristics of the level sensor needs to be
determined. The level sensor is a linear device, which may exhibit hysteresis in the output. The Introduction
output (voltage) can be related to the input (level) by the equation:
A control valve is a device which allows to change flow resistance, through a changing external
Level (%) = Ks ⋅Voltage(V ) + Zs (1) input signal. A control valve changes the flow of a fluid by altering the valve's flow resistance.

A control valve consists of two main components:

where K s is called the sensor’s Gain (defined as changed in the Output over change in the input)
- the valve actuator (which accepts an external signal and changes the valve stem position) and
and Zs is the Zero of the instrument. - the valve body (which contains a movable plug attached to the valve stem). By partially
blocking the valve orifice with the valve plug, a variable flow resistance is obtained.
Calibration Procedure
Valve Actuator
1. First, determine the range of operation of the sensor's input variable (i.e. the liquid level)
Most valve actuators (including those found in the Control Laboratory) are linear devices, altering
and output variable (i.e. the voltage sent to the controller). This is done by:
the valve stem position, x , in response to a change in an input signal, u :
a. Setting the controller on AUTO mode and changing the setpoint on the
computer (0 to 100%) until the water level has stabilized at the desired x = a ⋅u +b (2)
MAXIMUM level in the tank (e.g. just below the overflow in the tank), and
where a and b are constants. The stem position varies between x = 0 to x = x max . The
noting the value of the measured signals (signal in voltage and level in %).
maximum stem travel, x max , varies from actuator to actuator; as a consequence, a dimensionless
b. Applying a similar procedure but changing the setpoint on the computer until the
valve stem position, is defined by:
level stabilizes at the desired MINIMUM level in the tank (e.g. at or slightly above
the bottom of the tank) l = x / x max (3)
2. Starting from the liquid level and setpoint identified in step 1 above, change the setpoint such that
progressively to a number of values (approximately 10) to adequately cover the range of l = 0 at minimum flow rate (4)
operation identified above. When the system stabilized at the new setpoint RECORD the l = 1 at maximum flow rate (5)
liquid level (in %) which is the input to the sensor, and the measured signal (Voltage)
which is the output of the sensor. Measure the input and output variables at sufficiently The valve action determines the values of a and b in (2).
many points (approximately 10) between the extremes of operation.
3. To assess the extent of hysteresis calibrate the sensor using strictly increasing liquid levels Valve Action
and then reverse the process using falling liquid levels. This can be achieved easily by
setting the controller on a Manual mode. If the tank has the minimum level, close the Control valves have two main types of action. They either
control valve completely. Due to the constant input flow the level will increase steadily
until the maximum level is reached. When the maximum level is reached open the control ƒ Fail Open (FO), producing maximum flow at minimum signal, giving
valve completely and turn off the input flowrate. The level will decrease continuously.
Use this data to perform calibration. Since the values used in this procedure are not u max − u
l= (6)
steady state values (they change continuously) the procedure is called dynamic calibration. u max − u min

ZK Nagy 3 ZK Nagy 4
Instrumentation, Control & Industrial Practice Laboratory Manual Instrumentation, Control & Industrial Practice Laboratory Manual

ƒ or Fail Closed (FC), producing minimum flow at minimum signal, giving Installed Characteristics

u − u min When a valve is installed in a pipeline it becomes part of the flow circuit. Because of the flow
l= (7) resistances of the other components of the flow circuit, the pressure drop across such a valve is
u max − u min
no longer constant but varies with the fluid flow rate
where u min and umax are the minimum and maximum values of the external input signal. Usually The installed characteristic of a valve is the flow behaviour when the valve is part of a fluid circuit.
this signal is electric or pneumatic. In our case we can consider a range of 0-100%. For fully turbulent flow through a circuit consisting of a centrifugal pump in series with a control
valve and a fixed flow resistance, the installed characteristic of a linear valve is:
Valve Characteristics
l
Q= (9)
The valve body contains an orifice, through which fluid flows, and a movable plug, which (1 − α)l 2 + α
obstructs the flow. Valves are distinguished by their characteristic, which is the variation in the
fluid flow through the valve as he valve plug is moved. Differently shaped plugs give different where Q = q / q max is the fractional flow, and α is relative valve resistance. To describe the flow
valve characteristics. Motion of the valve plug is reported as changes in the fractional valve stem circuit only one parameter, α , is needed.
position, l .
Relative Control Valve Resistance
Inherent Characteristics
The relative (control) valve resistance is defined as the ratio of the energy dissipated by the valve
The inherent characteristic of a valve is given by the flow behaviour through the valve when the to the energy dissipated by the system, at maximum flow, i.e.,
pressure drop across the valve body is constant. For fully turbulent liquid flows, the inherent
characteristic is reported using the valve sizing coefficient, C v or K v , which is the flow of water ⎡ Energy dissipated by the valve ⎤
α=⎢ ⎥ (10)
at unit pressure drop across the valve body. An overview of the basic valve equations is given ⎢⎣ Energy dissipated by the system ⎥⎦ maximum flow
below.
Properly-sized valves have values of α between 1/3 and 2/3 .

Experimental procedure for control valve calibration

1. Select the controller on MANUAL mode and set the manual control input in the
software interface so that the valve is completely open.
2. Adjust the water supply hand valve below the rotameter in the water inlet line so that a
steady state operation with the maximum level is achieved (this can be achieved for
example by setting the inlet flow such that the water flows out the overflow even when
the valve is fully open). This guarantees that any effect of changes in liquid height in the
tank on flow rate through the control valve is eliminated from the calibration.
3. Determine the range of operation of the valve’s output variable (the flow rate of water)
that corresponds to the valve’s input variable (the signal from the controller, i.e. 0-100%).
4. SET and RECORD the input variable to the valve at a number of points (approximately
10) and MEASURE and RECORD the output variable (flow rate of water through the
valve) from the electronic flowmeter installed on the output line. Cover adequately the
range identified in the previous step.
The valves in the Control Laboratory have linear inherent characteristics: 5. To assess the extent of hysteresis perform the calibration process by strictly increasing the
signal to the valve and then reverse the process using strictly decreasing signals.
C v = C v,max ⋅ l (8)
6. Plot the valve characteristics (flow rate versus signal to valve in %) and comment on the
results.

ZK Nagy 5 ZK Nagy 6
Instrumentation, Control & Industrial Practice Laboratory Manual Instrumentation, Control & Industrial Practice Laboratory Manual

Determination of the relative control valve resistance

PART 2 - WEEK 2
1. From the calibration curve and practical observations determine whether the valve is a
FO or FC valve. III. Experimental procedure for the study of feedback control loop.
2. Based on the outcome use equation (6) or (7) to calculate the steam positions l .
The objectives of the exercises are to gain familiarity with feedback control, to appreciate the
3. Calculate the fractional flow Q = q / q max . importance of automatic control and controller tuning, and to gain qualitative appreciation of the
4. Equation (9) can be rearranged as follows: effect of each term in a three-term controller on the control performance.

Manual Control
(Q −2 − 1) = α(l −2 − 1) (11)
1. With the controller in AUTO mode define a set point (e.g. 50%) and wait until the
−2 −2
system stabilizes.
Plot (Q − 1) versus (l − 1) . Fit a straight line passing through the origin to the data
2. Turn the control mode to MANUAL, and change the setpoint (e.g. to 70%).
points. The slope of the straight line gives α . Comment on the value obtained. Properly
sized valves have a value between 1/3 and 2/3. 3. With the manual action buttons in the Process Diagram pane try to bring the level to the
new setpoint and maintain it at the new value.
4. Repeat the procedure at a different setpoint.
5. Record the data and analyse the observations. Remember that you can record the
measurements in a file by pressing the save button at the moment when you want the
recording to start (e.g. before you change the MANUAL operation mode and change the
setpoint). Alternatively (or simultaneously), you can transfer the data to Excel, by pressing
PLEASE NOTE the Send to Excel button after the experiment. You can also record and paste in a
After Part I and II are completed and before leaving the laboratory, shut-down the rig Word document the figure plotted by the software by right clicking on the figure widow
using the nominal shut-down procedure set out in the appendix. and selecting “Export Simplified Image” and then paste into a word document. In your
report discuss how easy is to maintain the setpoint? This is one of the simplest processes,
where you only had to control one single parameter! What is the control error in the case
of manual control, i.e. what is the maximum deviation from the setpoint?
6. In the final report present the figures showing the outcome of manual control together
with your comments.

Automatic control with heuristic tuning

The most common controller used in the process industries is the so-called three term controller,
or PID (proportional-integral-derivative) control. The controller calculates the error between the
measured signal and the desired value and changes to manipulated value (valve position) so that
the error is decreased. The amount of change in the manipulated value is a function of the error.
This function in the case of PID control has three main components.

ZK Nagy 7 ZK Nagy 8
Instrumentation, Control & Industrial Practice Laboratory Manual Instrumentation, Control & Industrial Practice Laboratory Manual

Procedure

1. Set the control mode to AUTO, and wait until the system stabilizes at the desired
setpoint value (e.g. 40%).
2. Test the performance of the default control parameters by changing the three controller
parameters and comparing how the controller performs in the case of different setpoint
changes. Observe how large the overshoot is, how fast the system reaches the new
setpoint or whether there is a steady state error, and if there is how large it is. Analyse the
form and speed (noise level, oscillations, etc.) of the control action together with the
variation of the water level.
3. in the Controller pane of the software, modify one of the three “PID parameters” at a
time. Test first the P only control by setting Ti=0 and Td=0. Test different gains
(minimum 3 different values). Next, fix Kc and vary Ti, but keep Td = 0, and finally
fixing Kc and Ti vary Td.
4. Record the used control parameters and plot the results in your report. Analyse the effect
of each term on the control performance.

Automatic tuning

The systematic determination of the controller parameters is called tuning procedure. There are
numerous ways to determine the controller parameters based on the criteria considered. The
software includes an “Automatic Tuning Wizard” which can help to determine the set of tuning
parameters.

1. With the control mode on AUTO, wait until the system stabilizes at the desired level.
2. start the “Automatic Tuning Wizard” from the Controller pane of the window by
pressing the “Autotune” button. Follow carefully the procedure. Make sure that you wait
enough when the procedure requires to gather information from the system. During the
procedure observe that the Autotuning wizard will introduce relay signals in the system to
understand it dynamic behaviour. The information is used to determine the control
parameters. At the end of the procedure the new parameters will automatically be saved
into the current PID parameters tag in the controller pane and will be used in the system.
Test the control performance by changing the setpoint and/or introducing disturbance by
changing the input flowrate. Record the controller parameters and the resulted figures.
3. Repeat the autotuning wizard and tests for different controller types. Tune P, PI and PID
The control performance depends strongly on the three parameters of the three term controller: control structures with Fast, Slow and Normal action. It is not necessary to tune all
ƒ Proportional Gain ( Kc ) combinations, select e.g. normal action and tune P, PI, and PID structures, and then with
the PI determine the parameters for the fast and slow response.
ƒ Integral Time ( τI )
4. With a fixed set of controller parameters test the control performance for different input
ƒ Derivative Time ( τD ) flowrates. When the system reaches a stable level increase or decrease the input flow
within the physical limits of the controllable system. This will act as a disturbance in the
system and the controller will open or close the valve accordingly to bring back the level

ZK Nagy 9 ZK Nagy 10
Instrumentation, Control & Industrial Practice Laboratory Manual Instrumentation, Control & Industrial Practice Laboratory Manual

to its setpoint value. Determine what the maximum input flowrate is, for which the APPENDIX A
controller is not able to stabilize the system anymore. Explain in your report what the
reason is and give an explanation how the maximum input flowrate, for which the Nominal Start-up Procedure
controller can stabilize the system relates to the output flowrate from the tank.
1. Turn off all hand valves.
Effect of controller sign an system stability 2. Turn on the compressed air at the ball valve on the supply line.
3. Turn on the computer.
1. With the best set of controller parameters change the sign of the controller gain Kc. 4. Turn on the electricity main for the instrumentation.
5. Run the control program.
2. Give a setpoint change and observe and RECORD the controller action and system’s
6. To avoid surges through the glass rotameter, slowly open the water supply hand valve.
behaviour.
7. Set an desired input flow rate. Recommended values:

Rig 1: 0.3 m3/h

Rig 2: 5 l/min
Rig 3: 380 l/h
Rig 4: 3 l/min
Rig 5: 220 l/h
Rig 6: 6 l/min

8. When enough water is in the tank, turn on the outlet pump at the electricity mains.
9. Make sure the controller is in AUTO mode and allow the controller to bring the liquid
level to steady state.
PLEASE NOTE
After Part I and II are completed and before leaving the laboratory, shut-down the rig
Nominal Shut-down Procedure
using the nominal shut-down procedure set out in the appendix.
1. Turn the controller OFF and open the control valve.
2. Turn off the water supply at the hand valve.
3. When the tank is empty, turn off the outlet pump at the electricity mains.
4. Transfer any stored data to a floppy diskette, memory stick, your network drive, or email
to yourself.
5. Turn off the computer at the electricity mains.
6. Turn off the instrumentation’s power supply at the electricity mains.
7. Turn off the compressed air at the ball valve in the supply line.
8. Place beakers under the drain cocks and drain the system.
9. Close all drain cocks.
10. Make sure that all electrical devices are switched off and that water is not dripping from
the mains water supply.

ZK Nagy 11 ZK Nagy 12
Instrumentation, Control & Industrial Practice Laboratory Manual Instrumentation, Control & Industrial Practice Laboratory Manual

APPENDIX B Marking guidelines

Laboratory Report
Organization
A group report is required from each team. The body of the report should be no more than 20 Are the report contents structured?
pages. You should include key figures, tables and calculations in the main body of the report but Is a table of contents present?
you may attach any extra data and peripheral information to appendices, which may be of any Are the pages numbered?
size. Is a summary present?
Are equations numbered?
Structure and Guidelines for the report:
P&ID
Summary
Is the P&ID correct?
Provide a paragraph of the main purpose of the report and a summary of the most important
Are the manual valves labelled?
findings.
Are the symbols correct?
Table of contents
Figures & Graphs
Introduction (0.5-1 page) Are the figures and graphs clear?
A short introduction related to the importance of understanding process dynamics and control Is the precision of the data reflected in the graphs?
should be provided. The introduction must contain several references. Use appropriate reference Is proper analysis and discussion provided?
style. Are the figures numbered, captions used and referenced properly in the text?
Theory (2-5 pages)
Summary
Explain in a few sentences the theoretical background related to: (1) how the sensor works, (2)
Does the summary reflect the experimental observations?
(2) calibration, (3) how the control valve works, (4) PID control. This part should be your own
Are the numerical values of the principal parameters tabulated?
description not copied from the lab manual, with references demonstrating your readings on the
topics.

Experimental Procedure (2-4 pages)

Provide a brief summary of the procedure. It is not necessary to copy the procedure from the
manual. You should describe briefly the methodology. A proper P&ID of the experimental rig
must be provided using standard symbols for instrumentation. It is important that all instruments
and elements of your rig are shown in the P&ID.

Results and Discussions (6-10 pages)

Show all results requested in the lab manual. This should contain as a minimum: results from
level calibration, analysis of hysteresis, valve calibration, analysis of hysteresis, calculation of the
relative control valve resistance, figures and discussion on manual control, results on heuristic
tuning, automatic tuning and effect of controller signs. Always compare/contrast and analyse
your results, do not only show figures with simple plots. Make sure that the controller tuning
parameters are indicated for different parts of the figure and you correlate the dynamic response
in your discussion with the changes in the parmeters. Main figures should be inserted in the text
(always number figures and use figure caption).

Conclusions (0.5 page)

A brief statement with the main results and conclusion of the report.

References
Use proper referencing style and make sure all references are cited in the report.

Appendixes

ZK Nagy 13 ZK Nagy 14
Instrumentation, Control & Industrial Practice Laboratory Manual Instrumentation, Control & Industrial Practice Laboratory Manual

APPENDIX C

Introduction to the software interface

The software interface allows you to interact with the experimental rig and record all data. The
main parts of the interface are shown below. There are three panes with different functionality:

ƒ Process Diagram
ƒ Controller
ƒ Configuration

In the first week you need the functionality of the “Process Diagram” pane only, whereas in the
second week you will work with both the “Process Diagram” and the “Controller” panes. The
configuration pane should normally not be used during the laboratory exercise.

Fig. 3. The “Controller” pane of the laboratory software

Saving your data

Making sure that your data is saved is your responsibility. The software offers several ways to
record you data.

1. - By pressing the save to file button, your data will be saved in a data file, which can
be opened by any standard text editor or Excel. Your data will be saved from the
moment you pressed the button onward, until you stop the data recording. No data
acquired before the moment you activate data recording will be saved in the file. When
you activate data recording the software will prompt you for a file name. You can select a
file name in your home folder (U drive). The data is recorded at every sampling time. The
order of the columns is as follows:

Time (s); Level (%); SP (%); MV (%); FlowRate; Level(V);

2. - By pressing the “Send to Excel” button, you can send the data from the memory of
Fig. 2. The “Process Diagram” pane of the laboratory software the software to an Excel workbook. This option works in the opposite way as the Save
button. By pressing the button data stored in the memory from the moment you started
the software until the moment the button was pressed will be sent to Excel.

ZK Nagy 15 ZK Nagy 16
Instrumentation, Control & Industrial Practice Laboratory Manual

3. You can also copy paste snapshots of the plot into your report. However to perform the
data analysis you need the values saved with method 1 or 2 above. For the qualitative
analysis part or for the appendix it is acceptable however to use the snapshots. This can
be obtained by right clicking on the figure widow and selecting “Export Simplified
Image…”, then paste it in Word, as shown in the figure below.

Fig. 4. Exporting snapshots of the plot to other applications

ZK Nagy 17